Can you create an Obama sweep in your state?

[muon2]

The more interesting challenge is to find states or significant areas that could be divided so that Obama sweeps in 2008 and loses them all in 2012. I did that for IL without Chicagoland as part of Antonio's alternate states.

Here's a cute gerrymander of your lopped-off IllinoiS. I divided it into 6 CDs using whole counties and keeping all 6 CDs within 1500 (0.2%) of the quota. I assumed that the Pubs had control after 2010, but they were willing to take some risks to make a play for all 6 seats. So all 6 CDs here voted for Obama over McCain, but Obama never gets over 52% (53% of the two-party vote) and all have a R PVI.

When I calculated the 2012 results, the gamble worked. In 2012 the favorite son effect wears off, so in the first election with this map Romney wins all 6 CDs with at least 52% of the two-party vote. The PVIs are calculated with the 2008-12 numbers.



IS-01:R+3 (Blue, Rockford-LaSalle)
IS-02:R+3 (Green, Rock Island-Bloomington)
IS-03:R+3 (Purple, Peoria-Quincy)
IS-04:R+6 (Red, Champaign-Decatur)
IS-05:R+5 (Yellow, Springfield-Edwardsville)
IS-06:R+5 (Teal, Belleville-Carbondale)

An equivalent goal could be a complete flip of all districts between 2004 and 2008 in any of the swing states that flipped R to D.

[muon2]

Here's my entry for IN. No munis/townships are split, and in Indianapolis the old townships are preserved. All deviations are under 300, and all CDs were won by Obama in 2008. The districts are between R+2 and R+4, and given the swing in IN they were all likely lost by Obama in 2012.



CD 1: 50.2-49.0
CD 2: 49.8-49.2
CD 3: 49.6-49.4
CD 4: 50.0-48.9
CD 5: 49.6-49.2
CD 6: 50.2-48.7
CD 7: 49.7-49.3
CD 8: 50.9-48.0
CD 9: 49.7-49.1

[muon2]

Here's a neat party trick.  Michigan, where:

a) all districts were won by Obama in 2008
b) the VRA is kept, with two BVAP-majority districts
c) the usual Michigan rules are basically hewn to- no districts are more than 1K off ideal population, no munis are split besides Detroit, no gratuitous county splits or even bridges.  Caveat: I did not necessarily reach to maximize whole-county groups, however.  Some extra hunting may be able to get one or two more such groups, but I've made enough of these maps for now.
d) except for the two VRA districts, all districts have a Republican average.

If one throws out the VRA it should be possible to make 14 Obama/Republican districts; whether doing so is compatible with Michigan Rules is not something I've tested, but I'm dubious- I assume that you need to send out a number of double-spanning ribbons from Detroit to Grand Rapids.







District 1: Obama 49.5% (McCain 48.7%); Dem 45.1%.  The UP and other northern parts. Whole-county.
District 2: Obama 50.2%, Dem 42.1%. Bay City, Midland, the western shore, etc.
District 3: Obama 52.3%, Dem 39.9%.  Muskegon, Grand Rapids, Holland.
District 4: Obama 54.5%, Dem 47.3%.  Saginaw, the Thumb, north of Flint.
District 5: Obama 57.5%, Dem 48.5%.  Flint, Livingston County, Ypsilanti, north of Lansing.
District 6: Obama 53.1%, Dem 44.0%.  Kalamazoo, Niles/Benton Harbor.
District 7: Obama 53.3%, Dem 41.9%.  Lansing, Grand Rapids burbs.
District 8: Obama 57.0%, Dem 47.3%.  Ann Arbor, Battle Creek, Jackson.
District 9: Obama 51.1%, Dem 41.3%.  Parts of Macomb and Oakland.
District 10: Obama 54.6%, Dem 47.4%.  St. Clair and the rest of Macomb.
District 11: Obama 55.2%, Dem 43.9%.  The bulk of Oakland.
District 12: Obama 54.8%, Dem 45.9%.  Western Wayne, Monroe, part of Lenawee.
District 13: Obama 78.9%, Dem 69.4%.  53.3% black (51.9% BVAP).  Eastern Detroit, the Grosse Pointes, Romulus, Southtowns.
District 14: Obama 80.9%, Dem 71.3%.  51.6% black (51.2% BVAP).  Western Detroit, Dearborn, bit of Oakland.

Nice job. I'd say that districts 5 and 8 have surprisingly high Obama numbers to generate a sub 50% Dem result. Based on the presidential result those would be D+3 or D+4 districts.

When I looked at DRA it says that the partisan numbers are from the 2006 Gov, AG and SoS. That year Granholm (D) won with 56.4%, Cox (R) took AG with 53.8%, and Land (R) took SoS with 56.2%. DRA must use a straight average of those races to have such a net GOP lean that Dems would only get 48% in a D+4 district.

[muon2]

Can someone help me with this program? I've always wanted to use it, but it always ends up freezing the browser I use. It says that Microsoft Silverlight is unresponsive.

I have had problems with every browser except IE. DRA is one of the only things I use IE for. DRA was written by a Microsoft engineer, so it makes sense that using all MS products gives me the best result.

[muon2]

Here's my IA entry. It wouldn't be right to chop counties for an IA plan, so I kept them all intact. The maximum population deviation is 334. The challenge was to get the votes as equal as possible while minimizing deviations. All four districts are D+1 or D+2.



CD 1 (+125) Obama 54.1%
CD 2 (+307) Obama 54.5%
CD 3 (-334) Obama 53.4%
CD 4 (-99) Obama 53.8%

[muon2]

Challenges: Kansas, Arizona, Nevada

You can't do a sweep where Obama lost. Do have other thoughts for AZ or KS?

[muon2]

Train, can I just say you're now on my utterly awesome poster list?

Now here's a challenge: OHIO.

For the superwin: follow the VRA.

MWWAAHAAHAHAHA!!

I've started to do it, in between other projects.  Preliminary results: possible but super incredibly ugly- think a Cincy district, a Dayton district, a Steve LaTourette district, and 13 spaghetti strips going NE-SW.

Follow the VRA as well?  Nope, that's beyond my pay grade (and possibly actually impossible).

A low end VRA district makes the sweep mathematically possible. An all-Cuyahoga district (no split munis) would be 47.2% BVAP (43.4% WVAP) and should meet the law. That leaves the rest of state voting 49.4% Obama to 48.9% McCain. Getting the rest of the districts in that range may still not be feasible, but ... .

[muon2]

Train, can I just say you're now on my utterly awesome poster list?

Now here's a challenge: OHIO.

For the superwin: follow the VRA.

MWWAAHAAHAHAHA!!

I've started to do it, in between other projects.  Preliminary results: possible but super incredibly ugly- think a Cincy district, a Dayton district, a Steve LaTourette district, and 13 spaghetti strips going NE-SW.

Follow the VRA as well?  Nope, that's beyond my pay grade (and possibly actually impossible).

A low end VRA district makes the sweep mathematically possible. An all-Cuyahoga district (no split munis) would be 47.2% BVAP (43.4% WVAP) and should meet the law. That leaves the rest of state voting 49.4% Obama to 48.9% McCain. Getting the rest of the districts in that range may still not be feasible, but ... .

... your challenge is answered.

All districts are within 500 of the quota, and only 2 are more than 302 from the quota. No munis are split within a county except for Columbus (and I believe I split no Columbus wards). The Cuyahoga district is as described above. All other districts carried for Obama in 2008 with a plurality of the vote. The margin ranges from 438 to 2824 votes, but a win is a win.

[muon2]

I remember doing this for Minnesota once so I'll probably do it again. It's quite easy really, MN-01 can remain intact, MN-07 and MN-08 swap precincts to get MN-07 an Obama win and in the metro you have to do some weird carving and split Minneapolis to get some exurbs in with urban seats and then turn the old Bachmann seat into a weird sort of St. Cloud to inner suburbs strip. Barely for Obama, but won nonetheless.

Can you do it without chopping Mpls or StPaul?

[muon2]

Obviously a sweep is impossible, but I was able to carve out a map with 12 Obama 2008 CDs (pretty sure with a bit more planning, a 13-1 is possible).



CD-1: 50.3% Obama, 48.9% McCain
CD-2: 51.2% Obama, 47.9% McCain
CD-3: 49.6% Obama, 49.5% McCain
CD-4: 50.2% Obama, 49.0% McCain
CD-5: 49.7% Obama, 49.5% McCain
CD-6: 49.8% Obama, 49.7% McCain
CD-7: 49.8% Obama, 49.6% McCain
CD-8: 49.8% Obama, 49.4% McCain
CD-9: 49.7% Obama, 49.5% McCain
CD-10: 73.5% McCain, 25.7% Obama
CD-11: 49.7% Obama, 49.7% McCain
CD-12: 66.2% McCain, 33.2% Obama
CD-13: 49.9% Obama, 49.6% McCain
CD-14: 51.4% Obama, 48.0% McCain

The point contiguity is unfortunate (eg. 7-10, 10-11). Is it possible to draw the districts with regular contiguity only?